Project Description

Key information

Project title: MUSCLES: Unravelling Liquid Metal Embrittlement Across Scales
Project in the Spotlight: N19010
Written by Viktoria Savran

 

As the automotive industry moves towards lighter and more sustainable vehicles, advanced high-strength steels (AHSS) are becoming increasingly important. These materials combine high strength with reduced weight, helping manufacturers improve crash safety while lowering material consumption and CO₂ emissions. Yet the introduction of these steels also brings major manufacturing challenges. One of the most critical is liquid metal embrittlement (LME), a phenomenon that can lead to sudden cracking during welding operations.

The M2i project N19010, Multi-Scale Assessment of Liquid Metal Embrittlement at Steel-Zinc Interfaces (MUSCLES), set out to better understand this complex problem. The project was funded by NWO through the Open Technologieprogramma (OTP) 2020 under project number 18972. Bringing together researchers from the University of Groningen, TU Delft and Tata Steel, the project combined atomistic modelling, advanced welding simulations and detailed microstructural characterization to investigate LME across multiple length scales. The consortium worked towards a common goal: understanding how and why zinc-coated advanced steels lose ductility under specific thermo-mechanical conditions.

The project united three complementary research tracks. Florian Brunner at the University of Groningen (RUG) focused on atomistic modelling and fracture simulations, while Virgínia Morete Barbosa Bertolo and later Arjun Sood (both at TU Delft) investigated crack initiation and propagation mechanisms in Zn-coated TWIP steels. At TU Delft, Gautham Mahadevan studied the influence of welding conditions and grain boundary behaviour during resistance spot welding. The project was led by Prof. Francesco Maresca (RUG) and co-supervisors Prof. Vera Popovich and Marcel Hermans, with additional support from Prof. Roumen Petrov and Prof. Leo Kestens. From the industrial side, Soheil Sabooni (Tata Steel) played an important role in connecting the scientific insights generated within MUSCLES to industrial welding challenges and future steel development.

Although liquid metal embrittlement (LME) has been known for decades, predicting its occurrence remains extremely difficult. During welding, molten zinc from the protective coating can penetrate the steel substrate and trigger cracking. The process is influenced simultaneously by temperature, local stresses, diffusion behaviour and microstructure, making LME a truly multi-scale phenomenon.

Within MUSCLES, the consortium combined advanced experimental techniques such as SEM, EBSD, TEM and EPMA with atomistic, meso- and macro-scale simulations and welding experiments under realistic industrial conditions. One of the most important findings was the central role played by grain boundary character. The studies consistently showed that random high-angle grain boundaries were highly susceptible to zinc-assisted cracking, while Σ3 twin boundaries demonstrated remarkable resistance to embrittlement.

These findings point towards grain boundary ductility engineering as a promising future strategy for reducing LME sensitivity. By tailoring the grain boundary network in advanced steels, Tata Steel may be able to improve resistance against crack initiation and propagation during welding operations.

A major strength of MUSCLES was the close integration between modelling and experiments. At the nanoscale, new machine-learning-based interatomic potentials for Fe-Zn systems were developed, enabling highly detailed simulations of fracture behaviour and crack propagation. These simulations provided important insight into how zinc diffusion and intermetallic phase formation weaken grain boundaries at the atomic scale.

Figure 1. LME behaviour within the weld.

At the same time, experimental studies revealed that zinc penetration often occurs before visible cracks appear, suggesting that localized diffusion and grain refinement are already active in the earliest stages of failure. The project further demonstrated how welding parameters strongly influence LME severity. In particular, hold-time optimisation during resistance spot welding showed clear potential as a practical mitigation strategy by controlling zinc solidification and residual stress development.

The project also strengthened the understanding of how crack initiation and crack propagation can be decoupled, helping researchers identify the individual mechanisms governing different stages of LME development. This creates an important basis for future predictive welding models and improved manufacturing strategies for advanced steels.

While MUSCLES focused primarily on automotive steels, the broader implications extend well beyond a single industrial sector. The insights generated within the project are also relevant for green steel production, advanced welding technologies, corrosion and cracking phenomena, and other metal systems affected by liquid metal embrittlement.

Another important outcome of the project is the demonstration of how multi-scale modelling and advanced characterization can support future digital manufacturing approaches. By linking atomistic simulations, microstructural analysis and finite element modelling, the consortium showed how predictive approaches may eventually help industry assess LME sensitivity before production even begins.

For Tata Steel, the MUSCLES project has delivered a substantially improved understanding of liquid metal embrittlement (LME) as a multi-scale, system-level phenomenon rather than an isolated material limitation. By integrating atomistic modelling, microstructural characterization, diffusion analysis, and welding experiments, the project has demonstrated that LME develops through a cascade of interdependent mechanisms. These include grain boundary susceptibility to Zn-induced weakening, the connectivity of grain boundary networks that enables crack propagation, the diffusion-driven transport of Zn to the crack tip, and the thermo-mechanical conditions imposed during welding. This holistic view is particularly valuable, as it clarifies why LME is difficult to predict and identifies multiple controlling factors across different length scales.

From an industrial perspective, these insights translate directly into both short- and long-term benefits for Tata Steel. In the short term, the project highlights the critical role of welding parameters, especially hold time, in controlling the temporal overlap between liquid Zn and tensile stresses, providing a practical and immediately applicable route to reduce LME susceptibility and improve manufacturing robustness. In the longer term, the findings support the design of advanced steels with enhanced weldability through grain boundary engineering and optimized microstructures, while also laying the foundation for predictive modelling approaches. Such developments will enable more reliable definition of processing windows, reduce reliance on empirical trial-and-error, and support faster development of new steel grades, ultimately strengthening Tata Steel’s competitiveness in high-performance and lightweight applications.

For industries working towards lighter vehicles, greener production routes and more robust manufacturing processes, that understanding may prove increasingly valuable in the years ahead.